memgraph · gitbuda · Oct 17, 2025 · Oct 16, 2025 · Oct 17, 2025 · Oct 17, 2025
@@ -19,6 +19,7 @@ export default {
   "degree_centrality": "degree_centrality",
   "distance_calculator": "distance_calculator",
   "elasticsearch_synchronization": "elasticsearch_synchronization",
+  "embeddings": "embeddings",
   "export_util": "export_util",
   "gnn_link_prediction": "gnn_link_prediction",
   "gnn_node_classification": "gnn_node_classification",

@@ -0,0 +1,135 @@
+---
+title: embeddings
+description: Calculate sentence embeddings on node strings using pytorch.
+---
+
+# embeddings
+
+import { Cards } from 'nextra/components'
+import GitHub from '/components/icons/GitHub'
+
+The embeddings module provides tools for calculating sentence embeddings on node strings using pytorch.
+
+<Cards>
+  <Cards.Card
+    icon={<GitHub />}
+    title="Source code"
+    href="https://github.com/memgraph/mage/blob/main/python/embeddings.py"
+  />
+</Cards>
+
+| Trait               | Value               |
+| ------------------- | ------------------- |
+| **Module type**     | algorithm           |
+| **Implementation**  | Python              |
+| **Parallelism**     | parallel            |
+
+
+## Procedures
+
+### `compute()`
+
+The procedure computes the sentence embeddings on the string properties of nodes. Embeddings are 
+created as a property of the nodes in the graph.
+
+{<h4 className="custom-header"> Input: </h4>}
+
+- `input_nodes: List[Vertex]` (**OPTIONAL**) ➡ The list of nodes to compute the embeddings for. If not provided, the embeddings are computed for all nodes in the graph.
+- `embedding_property: string` ➡ The name of the property to store the embeddings in. This property is `embedding` by default.
+- `excluded_properties: List[string]` ➡ The list of properties to exclude from the embeddings computation. This list is empty by default.
+- `model_name: string` ➡ The name of the model to use for the embeddings computation, buy default this module uses the `all-MiniLM-L6-v2` model provided by the `sentence-transformers` library.
+- `batch_size: int` ➡ The batch size to use for the embeddings computation. This is set to `2000` by default.
+- `chunk_size: int` ➡ The number of batches per "chunk". This is used when computing embeddings across multiple GPUs, as this has to be done by spawning multiple processes. Each spawned process computes the embeddings for a single chunk. This is set to 48 by default.
+- `device: string|int|List[string|int]` ➡ The device to use for the embeddings computation. This can be any of the following:
+    - `"cpu"` - Use CPU for computation.
+    - `"cuda"` or `"all"` - Use all available CUDA devices for computation.
+    - `"cuda:id"` - Use a specific CUDA device for computation.
+    - `id` - Use a specific device for computation.
+    - `[id1, id2, ...]` - Use a list of device ids for computation.
+    - `["cuda:id1", "cuda:id2", ...]` - Use a list of CUDA devices for computation.
+by default, the first device (`0`) is used.
+
+{<h4 className="custom-header"> Output: </h4>}
+
+- `success: bool` ➡ Whether the embeddings computation was successful.
+
+{<h4 className="custom-header"> Usage: </h4>}
+
+To compute the embeddings across the entire graph with the default parameters, use the following query: 
+
+```cypher
+CALL embeddings.compute()
+YIELD success;
+```
+
+To compute the embeddings for a specific list of nodes, use the following query:
+
+
+```cypher
+MATCH (n)
+WITH n ORDER BY id(n)
+LIMIT 5
+WITH collect(n) AS subset
+CALL embeddings.compute(subset)
+YIELD success;
+```
+
+To run the computation on specific device(s), use the following query:
+
+```cypher
+CALL embeddings.compute(
+    NULL,
+    "embedding",
+    NULL,
+    "all-MiniLM-L6-v2",
+    2000,
+    48,
+    "cuda:1"
+)
+YIELD success;
+```
+
+
+## Example
+
+Create the following graph:
+
+```cypher
+CREATE (n:Node {id: 1, Title: "Stilton", Description: "A stinky cheese from the UK"}),
+(n:Node {id: 2, Title: "Roquefort", Description: "A blue cheese from France"}),
+(n:Node {id: 3, Title: "Cheddar", Description: "A yellow cheese from the UK"}),
+(n:Node {id: 4, Title: "Gouda", Description: "A Dutch cheese"}),
+(n:Node {id: 5, Title: "Parmesan", Description: "An Italian cheese"}),
+(n:Node {id: 6, Title: "Red Leicester", Description: "The best cheese in the world"});
+```
+
+Run the following query to compute the embeddings:
+
+```cypher
+CALL embeddings.compute()
+YIELD success;
+
+MATCH (n) 
+WHERE n.embedding IS NOT NULL 
+RETURN n.Title, n.embedding;
+```
+
+Results:
+
+```plaintext
++---------+
+| success |
++---------+
+| true    |
++---------+
++----------------------------------------------------------------------+----------------------------------------------------------------------+
+| n.Title                                                              | n.embedding                                                          |
++----------------------------------------------------------------------+----------------------------------------------------------------------+
+| "Stilton"                                                            | [-0.0485366, -0.021823, 0.0159757, 0.0376443, 0.00594089, -0.0044... |
+| "Roquefort"                                                          | [-0.0252884, 0.0250485, -0.0249728, 0.0571037, 0.0386177, 0.03863... |
+| "Cheddar"                                                            | [-0.0129724, -0.00756301, -0.00379329, 0.0037531, -0.0134941, 0.0... |
+| "Gouda"                                                              | [0.0128716, 0.025435, -0.0288951, 0.0177759, -0.0624398, 0.043577... |
+| "Parmesan"                                                           | [-0.0755439, 0.00906182, -0.010977, 0.0208911, -0.0527448, 0.0085... |
+| "Red Leicester"                                                      | [-0.0244318, -0.0280038, -0.0373183, 0.0284436, -0.0277753, 0.066... |
++----------------------------------------------------------------------+----------------------------------------------------------------------+
+```
@@ -33,6 +33,9 @@ The following tags are available on Docker Hub:
 - `x.y` - production MAGE image
 - `x.y-relwithdebinfo` - contains debugging symbols and `gdb`
 - `x.y-malloc` - Memgraph compiled with `malloc`instead of `jemalloc` (x86_64 only)
+- `x.y-relwithdebinfo-cuda` - Memgraph built with CUDA support* - available since version `3.6.1`.
+
+*To run GPU-accelerated algorithms, you need to launch the container with the `--gpus all` flag.
 
 For versions prior to `3.2`, MAGE image tags included both MAGE and Memgraph versions, e.g.
 
@@ -125,9 +128,18 @@ curl https://sh.rustup.rs -sSf | sh -s -- -y
 export PATH="/root/.cargo/bin:${PATH}" 
 python3 -m pip install -r python/requirements.txt 
 python3 -m pip install -r cpp/memgraph/src/auth/reference_modules/requirements.txt
-python3 -m pip install torch-sparse torch-cluster torch-spline-conv torch-geometric torch-scatter -f https://data.pyg.org/whl/torch-2.3.0+cpu.html
-python3 -m pip install dgl -f https://data.dgl.ai/wheels/torch-2.3/repo.html
+python3 -m pip install torch-sparse torch-cluster torch-spline-conv torch-geometric torch-scatter -f https://data.pyg.org/whl/torch-2.6.0+cpu.html
+python3 -m pip install dgl -f https://data.dgl.ai/wheels/torch-2.6/repo.html
+```
+
+<Callout type="info">
+
+To install the dependencies for GPU-accelerated algorithms, you need to use the GPU-specific requirements file:
+
+```shell
+python3 -m pip install -r python/requirements-gpu.txt
 ```
+</Callout>
 
 {<h3 className="custom-header">Run the `setup` script</h3>}
 

@@ -622,7 +622,13 @@ CREATE (:Flight {AIR123: datetime("2021-04-21T14:15:00Z")});
 
 Maps for constructing `ZonedDateTime` values may have the following fields:
 `year`, `month`, `day`, `hour`, `minute`, `second`, `millisecond`,
-`microsecond` and `timezone`.
+`microsecond` and `timezone`. 
+- If some numeric fields are omitted, they default to the lowest value for that field:
+     - `0` for `year`, `hour`, `minute`, `second`, `millisecond`, and `microsecond`.
+     - `1` for `month` and `day`. 
+- If the `timezone` is omitted, it defaults to UTC. 
+- If all numeric fields are omitted and only the `timezone` is specified, then a `datetime`
+represents the **current time** in the that timezone.
 
 There are two options for the `timezone` field:
 * string: timezone name from the

@@ -42,6 +42,46 @@ troubleshoot in production.
 
 ## 🚀 Latest release
 
+### Memgraph v3.6.1 - October 16th, 2025
+
+{<h4 className="custom-header">⚠️ Breaking changes</h4>}
+
+- Changed the behaviour of calls like `datetime({timezone:"Europe/Brussels"})`.
+  Such calls return the current clock time. If you need the previous behaviour,
+  then explicitly state one or more numeric fields as `0`, e.g,
+  `datetime({year: 0, timezone:"Europe/Brussels"}`.
+  [#3339](https://github.com/memgraph/memgraph/pull/3339)
+
+{<h4 className="custom-header">🐞 Bug fixes</h4>}
+
+- Fixed `convert_c.to_tree()` memory leak. Using `to_tree` will not indefinably
+  increase the memory. [#3333](https://github.com/memgraph/memgraph/pull/3333)
+- Fixed an issue where `SHOW PRIVILEGES` ignored the target user's database
+  access rights. The `SHOW PRIVILEGES` now works correctly and will return an
+  empty result if the user does not have access to the target database.
+  [#3340](https://github.com/memgraph/memgraph/pull/3340)
+- Fixed a cache invalidation bug that would cause the IN LIST operator to
+  produce wrong results.
+  [#3342](https://github.com/memgraph/memgraph/pull/3342)
+- Fixed `datetime()` calls with only a `timezone` parameter (e.g.,
+  `datetime({timezone:"Europe/Brussels"})`) to return the current clock time,
+  instead of a time with all numeric fields set to their minimum value.
+  [#3339](https://github.com/memgraph/memgraph/pull/3339)
+
+### MAGE v3.6.1 - October 16th, 2025
+
+{<h4 className="custom-header">✨ New features</h4>}
+
+- Added `embedding` module, allowing users to compute sentence embeddings for
+  vertices on CPU or Nvidia GPU. 
+  [#662](https://github.com/memgraph/mage/pull/662).
+
+### Lab v3.6.0 - October 8th, 2025
+
+<LabReleasesClient version="3.6.0" />
+
+## Previous releases
+
 ### Memgraph v3.6.0 - October 8th, 2025
 
 {<h4 className="custom-header">⚠️ Breaking changes</h4>}
@@ -195,12 +235,6 @@ troubleshoot in production.
 - Fixed closing of database connections after migrations are completed.
   [#665](https://github.com/memgraph/mage/pull/665)
 
-### Lab v3.6.0 - October 8th, 2025
-
-<LabReleasesClient version="3.6.0" />
-
-## Previous releases
-
 ### Memgraph v3.5.2 - September 24th, 2025
 
 {<h4 className="custom-header">✨ New features</h4>}